Claude Opus 4.8 vs GPT-5.2
Anthropic vs OpenAI — benchmarks, pricing, and capabilities side by side.
- •GPT-5.2 has the higher intelligence index (74.4 vs 73.1)
- •GPT-5.2 is cheaper ($1.75 vs $6.30 per 1M input)
- •GPT-5.2 is faster
| Claude Opus 4.8 | GPT-5.2 | |
|---|---|---|
| Intelligence index | 73.1 | 74.4 |
| Developer | Anthropic | OpenAI |
| Type | LLM | LLM |
| Access | — | API only |
| Context window | — | 400,000 tokens |
| Input price | $6.30 / 1M | $1.75 / 1M |
| Output price | $25.00 / 1M | $14.00 / 1M |
| Speed | 66 tok/s | 73 tok/s |
| Released | May 28, 2026 | December 11, 2025 |
| Parameters | — | — |
| Input modalities | — | Text, Image |
| Output modalities | — | Text |
Shared benchmarks
Claude Opus 4.8
GPT-5.2
AA-LCR
67.7
72.7
GPQA Diamond
92
92.4
Humanity’s Last Exam
45.7
35.4
SciCode
53.5
52.1
Terminal-Bench
58.3
47
τ²-bench
94.4
84.8